One View Is Enough: In-the-Wild Monocular Pretraining for Novel View Generation
Monocular pretraining lifts single images into pseudo-target views via depth and reprojection, yielding OVIE, which rivals multi-view baselines at 116 FPS without inference-time depth or multi-view training pairs.
Published 2026Paris Poster Session 4 · Thu, Dec 10, 5:30 PM–7:30 PM local time · Paris Poster Hall▲ 5 on Hugging FaceCode ★ 82arXiv ↗OpenReview ↗
Only vote on papers you've read. Sign in with GitHub to vote.
Abstract
Monocular novel-view synthesis has long required multi-view image pairs for supervision, limiting training to a narrow set of purpose-built datasets. We propose in-the-wild monocular pretraining: a frozen depth estimator lifts each source image into 3D and reprojects under sampled poses to yield pseudo-target views; masked losses restrict supervision to valid regions and an adversarial objective covers disoccluded areas. Scaled to 30 million uncurated images, this produces OVIE, requiring only a source image and target pose at inference. Prior work trains without multi-view data but needs a depth estimator at inference, or drops this dependency but requires multi-view training pairs; OVIE is the first to require neither. Without multi-view supervision, OVIE rivals in-domain baselines on RealEstate10K and surpasses all on DL3DV, producing the most multi-view-consistent trajectories of any geometry-free method; brief multi-view fine-tuning outperforms all geometry-free methods on their training domain. At 116 FPS, it is over 600x faster than the fastest baseline. Code and pretrained models are at https://github.com/kyutai-labs/ovie; video results are on the project page, https://kyutai.org/blog/2026-04-14-ovie/.